These days, when you hear about college students and genetic AI, chances are you’re getting a glimpse of the debate surrounding the adoption of tools like ChatGPT. Is it help? (Yes! Great for research! Quick!) Or is it bad? (Boo! Wrong information! Cheating!). However, some startups see the arrival of genetic AI in the school environment as a positive and a foregone conclusion. And they build products to satisfy what they think will be a specific market opportunity.
Now one of them has raised some money to fulfill that ambition.
MagicSchool AI, which makes artificial intelligence generation tools for educational environments, closed a $15 million Series A round led by Bain Capital Ventures. Denver-based MagicSchool started with tools for educators, and founder and CEO Adeel Khan said in an interview that it now has about 4,000 teachers and schools using its products to design lessons, write tests and produce other educational materials. .
More recently, it has started making tools for students as well, provided through their schools. MagicSchool will use the funds to continue building more on both of these tracks, as well as work on attracting more customers, hiring talent and more.
This latest round also includes backing from some very notable investors. They include Adobe Ventures (whose parent Adobe is heavily involved with AI in its platform) and Common Sense Media (the age-based technology review specialist that has grown into productive AI with an AI guidelines partnership with OpenAI and chatbots reviews ). People in the round include Replit founder Amjad Masad, Clever co-founders Tyler Bosmeny and Rafael Garcia, and OutSchool co-founder Amir Nathoo. (Some of them were also early investors in the company: it had previously raised about $2.4 million.)
Khan did not disclose MagicSchool’s valuation in this round, but investors believe betting support for apps like this is the natural next step in AI startups after the hundreds of millions plowed into infrastructure companies like OpenAI, Anthropic and the Mistral.
“There’s an AI moment for education, a big opportunity to build an assistant for both teachers and students,” Christina Mela-Kyriazi, a partner at Bain Capital Ventures, said in an interview. “They have an opportunity here to help teachers with lesson planning and other tasks that take them away from their students.”
From teacher to AI preacher
MagicSchool, despite its name, was not created out of thin air.
Khan got his start as an educator, initially working for Teach for America when he first left university. (And his interest in public service and the role education plays may have started even before that: At Virginia Tech, he was student body president during his Virginia Tech shooting So, unfortunately, he had a front-row seat to the ravages of gun violence.)
As a teacher, he showed early signs of tapping into both business and leadership interests when he moved to Denver with the idea of starting a school of his own.
Working first in different administrative roles at local schools, he eventually founded his own, a charter high school called DSST: Conservatory Green High School, which subsequently saw its first group of graduates gain 100% acceptance into four-year colleges.
While taking a career break from this frenzy of activity, Khan came up with the idea for MagicSchool.
“It was around November 2022 when ChatGPT was dominating the headlines and genetic AI entered the ether for the majority of the country,” he recalls. “As I was thinking about what I was going to do next, I started thinking about it, and I immediately thought how much utility there was for educators in this new technology.”
He worked on early versions of using genetic artificial intelligence to build tools for teachers, visiting schools where he had taught and studying his former colleagues through the possibilities. But it didn’t click.
“The interface was awkward for them and it just wasn’t sticky,” he said. Khan’s demonstrations of them inspired the desired ‘wow’, but left to their own devices, teachers would use it once and never again.
“They’d say, ‘I spent so much time trying to motivate it and get it to do what I wanted, that it ended up not saving me time, but costing me time.’
His solution was to come up with more specific adjustments.
“Behind the scenes, we were just doing some very sophisticated prompts and also making sure the results were what an educator would expect,” he said.
Some of the examples of what teachers create with MagicSchool include lesson plans, quizzes and tests, course materials, and reformulations of prepared material for increasingly less challenging levels of learning. MagicSchool continues to deal with all of this. Khan said it works a lot with OpenAI’s APIs, but also with Anthropic and others. Behind the scenes, he said, the company does AB testing to determine what works best in which scenario.
However, getting teachers (who weren’t paying to use the product) and then schools (who were paying) to sign up for MagicSchool wasn’t exactly straightforward.
“I couldn’t meet with any schools or districts when we launched the product, including the one I was working at. There was a lot of fear about it all,” he said. All it took was “one negative headline about the use of AI in schools… about how AI will take over the world and robots” to end any discussion.
This gradually began to change as society and industry adopted AI more widely and more advanced models were released. Saving time was the most obvious reason for using it, he said, but they also found it was good for sharing ideas and even providing a supplement to what they could teach themselves.
“I think educators didn’t know or expect what AI could do for them and the public,” he said.
Plus, he has a second argument for why bringing more AI into the classroom makes sense: It’s going to be part of how everything is done, so it’s a school’s job to make sure its students are ready for it.
AI is smart but not ‘human smart’
That said, there are limitations to how AI can be used in any scenario, including the classroom.
“AI has a very different type of intelligence than human intelligence. Humans have developed emergent intelligence that is, in some ways, the product of millions of years of pruning through natural selection. It is very holistic. It’s very versatile, cognitively,” said Mutlu Cukurova, professor of education and artificial intelligence at University College London, where there is a multi-year research lab looking at the different permutations of artificial intelligence and learning. (A very realistic conclusion from a recent paper: There needs to be a hybrid approach involving both AI and humans.)
“AI is designed intelligence, not emergent intelligence. This means that it is designed for a very specific goal or set of goals. AIs are excellent at this particular goal and show significant signs of intelligence, but it’s a different type of intelligence.”
This may be particularly relevant to students and how they will learn in an AI world, or teachers who may not be experienced enough to know when an AI version of an educational material like a quiz is not good enough.
Cukurova said automating certain tasks can be a valuable use case, but “where it becomes problematic is when teachers … don’t have enough experience before they learn how to do these kinds of things on their own.”
Khan said the MagicSchool aims to address this especially when it comes to students. He said schools control what facilities they will provide to students on the platform and it is clear when they have used MagicSchool for an assignment.
All this sounds great in theory, but ultimately cracks can only be revealed in stress tests.
For example, will a cash-strapped school district rely more on in-class information from artificial intelligence systems with teachers? Or how will schools be able to recognize when students are using AI tools outside of the classroom in ways not approved by their teachers?
This will require a different kind of AI training, says Cukurova. “This is an important piece of the puzzle: How do we train and educate to use AI effectively and ethically?”
